Using current otoacoustic emission (OAE) screening protocols, approximately 50% of ears with mild hearing loss may be misidentified as having normal hearing. Currently accepted models of OAE generation suggest that there are two distinct mechanisms by which OAEs are generated in the cochlea. These include a nonlinear-distortion mechanism and a coherent-reflection mechanism. Distortion-product OAEs (DPOAEs) are believed to include contributions from both mechanisms, while stimulus-frequency OAEs (SFOAEs) are believe to include contributions primarily from the coherent-reflection mechanism. It has been proposed that OAEs may more accurately diagnose auditory pathology if the cochlear mechanisms contributing to the response are more carefully controlled. This program of research includes studies that will: (1) define the extent to which multiple-source interaction influences the accuracy with which DPOAEs identify hearing loss, (2) define the accuracy with which single-source OAEs (SFOAEs, DPOAEs with only the distortion-source contributing, and DPOAEs with only the reflection-source contributing) identify hearing loss, and (3) explore differential effects of stimulus parameters on DPOAE responses in normal and impaired ears with the goal of choosing stimuli based on their ability to distinguish the two populations. While the results obtained from these studies have direct implications for the accuracy with which we screen for hearing loss in infants and young children, they also have implications that extend beyond this population. These studies will define the extent to which distortion-source OAEs and reflection-source OAEs provide complementary information regarding cochlear function in large samples of normal-hearing and hearing-impaired ears. These data, therefore, will help to define the extent to which OAEs arising through different cochlear mechanisms can be used to characterize cochlear pathology.